Readynas OS6 TX - Uplink speed Problem // Slow Performance on iperf
Hello,
first, I have to retype everything because this forum deleted my first text. 😞
I have a serious performance Problem on my Readynas Ultra 2 with 2 GB and OS6. It seems Upload related. 1.1 Megabyte/s, while download (write) is 80,7 Megabyte/s. I did use the last stable version and now tried the new beta 2, but it didn't help...
I already tried a few solutions on this forum, but nothing helped. I use a Linux Machine as Client. Between me and my NAS are 2 Unifi Switsches (16 port with NAS --> 8 port with Client).
Iperf shows
Client iperf -c // NAS iperf -s (Client --> 16 Port --> 8 Port --> NAS)
[ 3] 0.0-10.0 sec 1.09 GBytes 935 Mbits/sec
Client iperf -s // NAS iperf -c
[ 4] 0.0-10.3 sec 22.1 MBytes 18.1 Mbits/sec
---
---
Retest with Raspberry ( RPI1 --> 16 port --> NAS) --> Seems OK (old RPI and slow 100MBit NIC)
RPI iperf -s // NAS iperf -c
[ 3] 0.0-10.0 sec 91.0 MBytes 76.0 Mbits/sec
RPI Iperf -c // NAS iperf -s
[ 5] 0.0-10.0 sec 48.9 MBytes 40.9 Mbits/sec
---
---
Linux Client Iperf to RPI showed similar results as test above.
---
---
Client to Client on 8-Port showed 935Mbit/s --> OK
---
---
Now to an interesting result
Linux Client --> NAS Iperf multiple Connections iperf -c 192.168.1.50 -P 10 ------------------------------------------------------------ Client connecting to 192.168.1.50, TCP port 5001 TCP window size: 85.0 KByte (default) ------------------------------------------------------------ [ 12] local 192.168.1.100 port 56586 connected with 192.168.1.50 port 5001 [ 3] local 192.168.1.100 port 56568 connected with 192.168.1.50 port 5001 [ 4] local 192.168.1.100 port 56570 connected with 192.168.1.50 port 5001 [ 5] local 192.168.1.100 port 56572 connected with 192.168.1.50 port 5001 [ 6] local 192.168.1.100 port 56574 connected with 192.168.1.50 port 5001 [ 8] local 192.168.1.100 port 56578 connected with 192.168.1.50 port 5001 [ 9] local 192.168.1.100 port 56580 connected with 192.168.1.50 port 5001 [ 7] local 192.168.1.100 port 56576 connected with 192.168.1.50 port 5001 [ 10] local 192.168.1.100 port 56582 connected with 192.168.1.50 port 5001 [ 11] local 192.168.1.100 port 56584 connected with 192.168.1.50 port 5001 [ ID] Interval Transfer Bandwidth [ 4] 0.0-10.0 sec 113 MBytes 94.7 Mbits/sec [ 8] 0.0-10.0 sec 113 MBytes 94.8 Mbits/sec [ 3] 0.0-10.0 sec 113 MBytes 94.8 Mbits/sec [ 9] 0.0-10.0 sec 113 MBytes 94.7 Mbits/sec [ 12] 0.0-10.0 sec 113 MBytes 94.8 Mbits/sec [ 6] 0.0-10.0 sec 113 MBytes 94.7 Mbits/sec [ 7] 0.0-10.0 sec 113 MBytes 94.7 Mbits/sec [ 5] 0.0-10.0 sec 113 MBytes 94.7 Mbits/sec [ 10] 0.0-10.0 sec 113 MBytes 94.5 Mbits/sec [ 11] 0.0-10.0 sec 113 MBytes 94.7 Mbits/sec [SUM] 0.0-10.0 sec 1.10 GBytes 945 Mbits/sec --> Looks like 1 Gbit/s
---
---
Tests showed, that I need a min of 7 simultanous Connections to get 1 Gbit --> But I dont thing that my NAS is doing that on HDDs
root@NetgearNAS:~# iperf -c 192.168.1.50 -P 2 ------------------------------------------------------------ Client connecting to 192.168.1.50, TCP port 5001 TCP window size: 85.0 KByte (default) ------------------------------------------------------------ [ 4] local 192.168.1.100 port 56590 connected with 192.168.1.50 port 5001 [ 3] local 192.168.1.100 port 56588 connected with 192.168.1.50 port 5001 [ ID] Interval Transfer Bandwidth [ 3] 0.0-10.1 sec 10.8 MBytes 8.90 Mbits/sec [ 4] 0.0-10.3 sec 9.38 MBytes 7.63 Mbits/sec [SUM] 0.0-10.3 sec 20.1 MBytes 16.4 Mbits/sec root@NetgearNAS:~# iperf -c 192.168.1.50 -P 4 ------------------------------------------------------------ Client connecting to 192.168.1.50, TCP port 5001 TCP window size: 85.0 KByte (default) ------------------------------------------------------------ [ 6] local 192.168.1.100 port 56598 connected with 192.168.1.50 port 5001 [ 4] local 192.168.1.100 port 56592 connected with 192.168.1.50 port 5001 [ 3] local 192.168.1.100 port 56594 connected with 192.168.1.50 port 5001 [ 5] local 192.168.1.100 port 56596 connected with 192.168.1.50 port 5001 [ ID] Interval Transfer Bandwidth [ 4] 0.0-10.4 sec 8.62 MBytes 6.98 Mbits/sec [ 3] 0.0-10.6 sec 5.12 MBytes 4.07 Mbits/sec [ 6] 0.0-10.8 sec 5.25 MBytes 4.09 Mbits/sec [ 5] 0.0-11.2 sec 5.12 MBytes 3.85 Mbits/sec [SUM] 0.0-11.2 sec 24.1 MBytes 18.1 Mbits/sec root@NetgearNAS:~# iperf -c 192.168.1.50 -P 6 ------------------------------------------------------------ Client connecting to 192.168.1.50, TCP port 5001 TCP window size: 85.0 KByte (default) ------------------------------------------------------------ [ 8] local 192.168.1.100 port 56610 connected with 192.168.1.50 port 5001 [ 4] local 192.168.1.100 port 56602 connected with 192.168.1.50 port 5001 [ 7] local 192.168.1.100 port 56608 connected with 192.168.1.50 port 5001 [ 6] local 192.168.1.100 port 56606 connected with 192.168.1.50 port 5001 [ 3] local 192.168.1.100 port 56600 connected with 192.168.1.50 port 5001 [ 5] local 192.168.1.100 port 56604 connected with 192.168.1.50 port 5001 [ ID] Interval Transfer Bandwidth [ 7] 0.0-10.0 sec 4.75 MBytes 3.97 Mbits/sec [ 4] 0.0-10.4 sec 6.75 MBytes 5.44 Mbits/sec [ 5] 0.0-10.4 sec 4.38 MBytes 3.52 Mbits/sec [ 6] 0.0-10.5 sec 4.50 MBytes 3.61 Mbits/sec [ 8] 0.0-10.6 sec 5.12 MBytes 4.07 Mbits/sec [ 3] 0.0-10.8 sec 4.75 MBytes 3.68 Mbits/sec [SUM] 0.0-10.8 sec 30.2 MBytes 23.4 Mbits/sec root@NetgearNAS:~# iperf -c 192.168.1.50 -P 8 ------------------------------------------------------------ Client connecting to 192.168.1.50, TCP port 5001 TCP window size: 85.0 KByte (default) ------------------------------------------------------------ [ 10] local 192.168.1.100 port 56626 connected with 192.168.1.50 port 5001 [ 4] local 192.168.1.100 port 56614 connected with 192.168.1.50 port 5001 [ 6] local 192.168.1.100 port 56618 connected with 192.168.1.50 port 5001 [ 7] local 192.168.1.100 port 56620 connected with 192.168.1.50 port 5001 [ 5] local 192.168.1.100 port 56616 connected with 192.168.1.50 port 5001 [ 8] local 192.168.1.100 port 56622 connected with 192.168.1.50 port 5001 [ 3] local 192.168.1.100 port 56612 connected with 192.168.1.50 port 5001 [ 9] local 192.168.1.100 port 56624 connected with 192.168.1.50 port 5001 [ ID] Interval Transfer Bandwidth [ 10] 0.0-10.0 sec 142 MBytes 119 Mbits/sec [ 4] 0.0-10.0 sec 138 MBytes 116 Mbits/sec [ 3] 0.0-10.0 sec 143 MBytes 120 Mbits/sec [ 7] 0.0-10.0 sec 139 MBytes 116 Mbits/sec [ 5] 0.0-10.0 sec 142 MBytes 119 Mbits/sec [ 9] 0.0-10.0 sec 144 MBytes 121 Mbits/sec [ 6] 0.0-10.0 sec 140 MBytes 117 Mbits/sec [ 8] 0.0-10.0 sec 142 MBytes 119 Mbits/sec [SUM] 0.0-10.0 sec 1.10 GBytes 945 Mbits/sec
---
So does anyone have a clue? Antivirus/Indexing/IPv6/SMB encryption/apps/ are turned of
Also switched NAS and Switch Ports and Cables.
Thx in advance
BR
PS First text was longer....
Model: ReadyNAS RNDU2000|ReadyNAS Ultra 2 Chassis only
Re: Readynas OS6 TX - Uplink speed Problem // Slow Performance on iperf
Hi, thx for your reply.
After a few ours of fiddling around, I just found the answer. It wasnt the Readynas, so this is the wrong forum. Sorry for that.
But you were on the right way. The 8 Port unifi Switch had a problem, flow control was turned off. So I turned it on just to see what happens. Results were the same, so I turned it off again and suddenly I got the desired performance of 945 Mbit.
I now can connect my Imac to my Nas with 84/102 Mbyte/s over afp. But on SMB I only get 30/30....
I will check this later...
Thank you I think we can close this. Flow Control needs to be turned off, even when the admin panel showed that it was off.
Not necessarily, but adjusting flow control should be part of the troubleshooting.
The original ethernet flow control is actually a "pause". So if too much traffic is going to a device, flow control will block all newtraffic going to the device until the queue drains. No packets would be lost.
If ethernet flow control is disabled, then the queue will overflow and packets will be lost. TCP's congestion control will back off the rate significantly in response to the packet loss, and it will slowly increase the rate that. There are several TCP congestion control algorithms in the wild - it's been enhanced/tuned over the years. So different senders will behave differently when there is packet loss.
In my own experience, enabling ethernet flow control end-to-end (switches and clients) often helps performance, especially if you have some older devices on slower links. But this isn't a "one size fits all" thing - so you basically need to try it both ways (similar to Jumbo Frames).
